INDEX

Explanations

references to political figures with specific names

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals, particularly those involved in politics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fertile

-0.70

 brightest

-0.68

matically

-0.67

 indo

-0.66

FAA

-0.66

################

-0.62

direction

-0.60

 takeoff

-0.60

 Flying

-0.60

 latitude

-0.59

POSITIVE LOGITS

otte

1.58

borg

1.07

reau

0.90

sburg

0.89

boro

0.85

quist

0.83

ecake

0.82

nil

0.82

yre

0.80

etz

0.79

Activations Density 0.003%

references to political figures with specific names

mentions of specific individuals, particularly those involved in politics

No Comments

No Known Activations